Security
Security and data-protection measures of the AI Waiter platform.
Updated: 17.07.2026
Our approach to security
Protecting customer and guest data is a core requirement of the platform. We apply layered controls across infrastructure, application code and operations.
Technical measures
Encryption
- All connections are protected with TLS 1.2+.
- AI provider API keys and messenger tokens are stored encrypted at rest.
- User passwords are hashed with bcrypt.
Authentication and authorisation
- Role-based access control (super-admin / restaurant staff / guest).
- Strict tenant isolation: every restaurant's data is scoped at the application layer, and panel queries cannot cross tenants.
- CSRF protection on all forms; output escaping against XSS, including the embeddable widget.
- Rate limiting on public API endpoints and AI-assisted features.
Operations
- Regular automated database backups.
- Availability and anomaly monitoring.
- Dependency updates and security patches applied promptly after publication.
- Configurable retention and automatic pruning of chat logs.
Responsible disclosure
If you discover a vulnerability, please report it privately before any public disclosure. We commit to:
- acknowledging your report within 48 hours;
- providing progress updates within 7 days;
- crediting researchers (if desired) after the fix is released.
Report to: [email protected]
Please do not run destructive tests (DoS, data deletion) and do not access other customers' data.
Compliance
- Personal data is processed in accordance with the GDPR and the Law of Ukraine «On Personal Data Protection».
- Payment card data is handled entirely by our payment partner Paddle (PCI DSS compliant); we never store card numbers.
- Data is not shared with third parties except as described in the Privacy Policy.
Last updated: 12 June 2026